A ketone complex by alkylation of an acyl anion. Synthesis, crystal structure and spectroscopic characterization of [Cp(CO)2Re{OC(Me)Ph}]

Article abstract of DOI:10.1016/S0020-1693(02)01520-7

Reaction of Li[Cp(CO)₂Re(COPh)] (1) with CF₃SO ₃Me afforded, besides the expected carbene complex [Cp(CO) ₂Re=C(OMe)(Ph)] (2) and the alkyl-acyl derivative [Cp(CO) ₂Re(Me)(COPh)] (3), a third structural isomer [Cp(CO) ₂Re{OC(Me)Ph}] (4), which contains an acetophenone molecule coordinated to the metal center. The X-ray analysis showed that in solid acetophenone is bound to 'CpRe(CO)₂' exclusively through an oxygen σ-donor interaction, while in solution an equilibrium between σ-bound (η¹) and π-bound (η²) forms occurs, as judged by IR data and ¹H and ¹³C variable temperature NMR spectra (π/σ ratio 2.87 at 183 K and 1.16 at 263 K in CD₂Cl₂, ΔH°=-4.5 kJ mol^-1 for the σ?π reaction, E_a 58(1) kJ mol^-1). In solvents different from Et₂O (n-hexane, THF, acetone) and with alkylating agent different from CF₃SO₃Me (MeI, Me ₃OBF₄) the formation of 4 was negligible. It has been demonstrated that 4 does not originate by acetophenone reductive elimination from 3. No evidence of the involvement of radicals has been obtained.

Full text of DOI:10.1016/S0020-1693(02)01520-7

Inorganica Chimica Acta 350 (2003) 475ꢀ485
/
www.elsevier.com/locate/ica
A ketone complex by alkylation of an acyl anion. Synthesis, crystal
structure and spectroscopic characterization of
[Cp(CO)2Re{OC(Me)Ph}]
M. Bergamo a, T. Beringhelli a, G. D’Alfonso a,*, D. Maggioni a, P. Mercandelli b,*,
A. Sironi b
a Dipartimento di Chimica Inorganica, Metallorganica ed Analitica, via Venezian 21, 20133 Milan, Italy
b Dipartimento di Chimica Strutturale e Stereochimica Inorganica, via Venezian 21, 20133 Milan, Italy
Received 15 July 2002; accepted 10 October 2002
Dedicated in honor of Professor Pierre Braunstein
Abstract
Reaction of Li[Cp(CO)2Re(COPh)] (1) with CF3SO3Me afforded, besides the expected carbene complex [Cp(CO)2Reꢀ
/
C(OMe)(Ph)] (2) and the alkylꢀacyl derivative [Cp(CO)2Re(Me)(COPh)] (3), a third structural isomer [Cp(CO)2Re{OC(Me)Ph}]
/
(4), which contains an acetophenone molecule coordinated to the metal center. The X-ray analysis showed that in solid
acetophenone is bound to ‘CpRe(CO)2’ exclusively through an oxygen s-donor interaction, while in solution an equilibrium
between s-bound (h1) and p-bound (h2) forms occurs, as judged by IR data and 1H and 13C variable temperature NMR spectra (p/
s ratio 2.87 at 183 K and 1.16 at 263 K in CD2Cl2, DH8ꢁ
/
ꢂ
/
4.5 kJ molꢂ1 for the sUp reaction, Ea 58(1) kJ molꢂ1). In solvents
/
different from Et2O (n-hexane, THF, acetone) and with alkylating agent different from CF3SO3Me (MeI, Me3OBF4) the formation
of 4 was negligible. It has been demonstrated that 4 does not originate by acetophenone reductive elimination from 3. No evidence
of the involvement of radicals has been obtained.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Ketone complexes; Acylmetallate; Alkylating agents; Methyl triflate; sꢀp Equilibrium; Rhenium
/
1. Introduction
[Cp(CO)2Re{m-CPh(OMe)}Pt(COD)] (CODꢁ
clooctadiene) [1g], was obtained by addition of a
nucleophilic Pt(COD) fragment over the ReꢀC double
C(OMe)-
(Ph)] [2], according to the approach developed by the
Stone group [3].
/
1,5-cy-
We report here on the unexpected formation of a
ketone complex in the reaction of an alkylating agent
with an acylrhenate salt. We met with this reaction in
/
bond of the Fischer’s carbene [Cp(CO)2Reꢀ
/
the course of our work aimed at preparing novel PtÃ
/Re
The original synthesis [2] of [Cp(CO)2Reꢀ
/
mixed-metal complexes. We are particularly interested
in complexes that do not contain p-donor ligands, since
these should be better suitable for catalytic trials, and in
the past we prepared several of these species [1]. The last
member of this series, namely the dinuclear complex
C(OMe)(Ph)] involved protonation of the acylmetalate
intermediate, followed by treatment with CH2N2, as
shown in Eqs. (1)ꢀ(3):
/
[CpRe(CO)3]ꢃLiR 0 Liꢃ[Cp(CO)2ReC(O)R]ꢂ
(1)
[Cp(CO)2ReC(O)R]ꢂLiꢃ ꢃHꢃ
0 [Cp(CO)2ReC(OH)R]ꢃLiꢃ
[Cp(CO)2ReC(OH)R]ꢃCH2N2
(2)
(3)
* Corresponding authors. Tel.: ꢃ
14405 (G.D.A.); tel.: ꢃ39-02-503 14447; fax: ꢃ
(P.M.).
/
39-02-503 14351; fax: ꢃ
/
39-02-503
/
/
39-02-503 14454
0 [Cp(CO)2ReꢀC(OMe)R]ꢃN2
E-mail addresses: giuseppe.dalfonso@unimi.it (G. D’Alfonso),
pierluigi.mercandelli@unimi.it (P. Mercandelli).
To prepare [Cp(CO)2Reꢀ/C(OMe)(Ph)] we replaced
0020-1693/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0020-1693(02)01520-7

476
M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ485
/
the above reported three-step procedure by the more
direct pathway nowadays used for the synthesis of this
type of carbenes, that consists in the direct alkylation of
a salt of the acylmetalate anion [4]. We, therefore,
treated Li[Cp(CO)2ReC(O)Ph] (Liꢃ1, formed according
to reaction 1) with the powerful methylating agent
CF3SO3Me, in diethyl ether solution. A thorough study
[5] of the alkylation of an acylrhenate closely related to
1, namely [Cp(CO)2ReC(O)Me]ꢂ, has previously de-
monstrated that alkylation can occur either at the
oxygen atom of the acylate group or at the metal center,
leading to a carbene or to an alkylꢀacyl derivative,
/
respectively. The site of alkylation has been shown to
depend on the hardness of the alkylating agent and on
the solvent, harder electrophiles and less polar aprotic
solvents favoring the formation of the carbene deriva-
tives. Neither our alkylating agent (methyl triflate) nor
our solvent (diethyl ether, Et2O) was used in Bergman’s
study, that rather used R3OPF6, RI, MeOSO2C6H4Me
(RꢁMe or Et) in acetone, THF or water. Under our
/
conditions we have observed the formation not only of
the products of the two above described pathways, i.e.
the carbene [Cp(CO)2Reꢀ
/
{C(OMe)Ph}] (2) and the
methylꢀbenzoyl derivative [Cp(CO)2Re(Me){C(O)Ph}]
/
Fig. 1. IR spectra (carbonyl region, diethyl ether) of a typical mixture
from the reaction of Liꢃ1 with CF3SO3Me (top trace) and of the three
main compounds isolated by chromatography from this mixture.
(3), but also of a third isomer, that contains a
coordinated acetophenone molecule and that does not
originates from 3 (Scheme 1). The ketone molecule is s-
bound in the solid state, while in solution an equilibrium
between the s and p forms of the ketone complex
occurs, whose thermodynamics and kinetics has been
investigated. To the best of our knowledge, this repre-
sents the first report of a ketone complex obtained
through a similar route.
phenylmethoxy carbene 2, even if its spectroscopic
data are somewhat different from those reported in the
1
original paper: H NMR, C6D6, d 4.65 (Cp), 4.16 (Me)
ppm (lit. 4.92, 4.19 in the same solvent); IR n(CO) (nujol
mull): 1960sh, 1940vs, 1888sh, 1872vs cmꢂ1 (lit. 1976ss,
1892ss cmꢂ1). Samples of compound 2 synthesized in
this way were successfully used for the synthesis of
[Cp(CO)2Re{m-CPh(OMe)2}Pt(COD)] [1g], and, there-
fore, its formulation should be considered correct.
The third eluted product was identified spectroscopi-
cally (see Table 1) as the complex [Cp(CO)2R-
e(Me){C(O)Ph}] (3), arising from alkylation on the
rhenium atom. Particularly diagnostic for this formula-
tion are the high filed resonance of the methyl group (d
0.85 ppm in C6D6) and the IR data [n(CO) in C6H6:
2007m, 1933s, 1603w cmꢂ1], strictly comparable with
those of the analogous methyl derivative [Cp(CO)2R-
e(Me){C(O)Me}], whose X-ray analysis showed a four-
legged piano stool structure with a trans arrangement of
2. Results and discussion
The reaction of Li[Cp(CO)2ReC(O)Ph] (Liꢃ1) with a
stoichiometric amount of CF3SO3Me, at room tempera-
ture, in diethyl ether, went to completeness in approxi-
mately 2 h, affording a deep red solution, from which
three main compounds were separated by chromato-
graphy on silica gel. On using a 3:1 n-hexaneꢀdiethyl
/
ether mixture, three bands (yellow, red, and light-brown,
respectively) were successively eluted (see Fig. 1). The
first product was identified as the expected yellow
the acyl and alkyl ligands (2004m, 1932s, 1630m cmꢂ1
in C6D6) [5].
The red product 4 eluted between 2 and 3 has NMR
,
(see Table 1), mass spectra data (FAB, molecular peaks
428ꢀ426) and elemental analysis that suggest its for-
/
mulation as an isomer of the two above species. This
was confirmed by a single crystal X-ray analysis, that
revealed the presence of an acetophenone molecule, s-
coordinated through the oxygen atom (Fig. 2).
Scheme 1.

M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ
/485
477
Table 1
Table 2
1
a
˚
IR and H NMR data for the compounds described in this work (298
K)
Selected bond lengths (A) and angles (8) for compound 4
Bond lengths
Compound nCO
1H NMR
Ph
Re(1)Ã
Re(1)Ã
Re(1)Ã
Re(1)Ã
/
C(1)
C(2)
cp(1)
O(3)
1.877(6)
1.866(6)
1.938(4)
2.100(4)
1.153(7)
1.176(8)
1.239(7)
/
Cp
Me
/
/
1
2
1918s, 1801s a
1958s, 1885s a
1964s b, 1893s
7.19ꢀ
7.30ꢀ
7.32ꢀ
/
7.70m g 4.92s
-
C(1)Ã
C(2)Ã
O(3)Ã
/
O(1)
/
6.88 d
6.99 f
4.65s
5.31s
4.17s
4.29s
0.85s
1.18s
2.33s
2.49s
/
O(2)
/
/
C(8)
3
4
5
2010m, 1940vs, 1608 vw a 8.00ꢀ
/
7.88m d 4.53s
7.58m e 5.62s
6.95m d 4.60s
7.33m e 5.26s
2014m, 1946vs, 1614 vw b 7.74ꢀ
/
Bond angles
C(1)ÃRe(1)Ã
C(1)ÃRe(1)Ã
C(1)ÃRe(1)Ã
C(2)ÃRe(1)Ã
C(2)ÃRe(1)Ã
cp(1)ÃRe(1)Ã
Re(1)Ã
Re(1)Ã
Re(1)Ã
O(3)ÃC(8)Ã
O(3)ÃC(8)Ã
C(9)ÃC(8)Ã
979m, 1909s, 1851w a
1985m, 1916s, 1861wb
2042s, 1959s c
7.35ꢀ
7.47ꢀ
/
/
/
C(2)
cp(1)
O(3)
cp(1)
O(3)
88.8(3)
123.3(3)
97.6(2)
/
/
/
5.85s e,h
/
/
/
/
124.2(2)
99.1(2)
a
b
c
/
/
Diethyl ether.
n-Hexane.
CH2Cl2.
/
/
O(3)
117.27(16)
173.3(7)
172.4(6)
139.0(4)
119.6(6)
117.6(5)
122.8(6)
/
C(1)Ã
C(2)Ã
O(3)Ã
C(9)
C(10)
C(10)
/
O(1)
d
e
/
/
O(2)
Benzene-d6.
CD2Cl2.
/
/
C(8)
f
/
/
Acetone-d6.
THF-d8.
The protons of the CH2Cl group give an AB system at d 4.72 and
g
h
/
/
/
/
2
4.49, with JHÃH
ꢁ/9 Hz.
a
cp(1) Refers to the centroid of the cyclopentadienyl ligand.
2.1. Solid state molecular structure of (acetophenone-
kO)dicarbonyl(h5-cyclopentadienyl)rhenium (4)
different steric demand of the methyl and phenyl
groups, the E isomer is the one observed.
˚
O(3) bond length (1.239(7) A)
As expected, the C(8)ꢀ
/
The crystals of 4 contain neutral molecules held
together by normal van der Waals interactions. Fig. 2
depicts the molecular architecture of 4 with a labelling
scheme. A selection of relevant geometrical bond para-
meters is reported in Table 2.
is slightly longer than that of free acetophenone
˚
(1.216(2) A) [6], as a consequence of the s bonding
and p back-bonding interactions with the metal center,
but substantially identical to that observed in the related
[CpRe(NO)(PPh3)(MeC(O)Ph)]ꢃ complex (1.245(8) A)
˚
The molecule, of idealized Cs symmetry, consists of a
rhenium atom pseudo-tetrahedrally coordinated by two
carbonyls, a cyclopentadienyl and a s-acetophenone
ligand. Unsymmetrical ketone complexes can exist as
[7]. The Re(1)Ã
/
O(3)ÃC(8) bond angle, 139.0(4)8 are
/
somewhat greater than it would be expected for a
bonding model involving an sp2-hybridized oxygen
atom. However, this feature is found in nearly all s-
ketone complexes of transition metals, and has been
either E or Z Cꢀ
/
O geometric isomers. Due to the very
Fig. 2. ORTEP drawing of molecule 4 with labeling scheme. Hydrogen atoms were given arbitrary radii. Thermal displacement parameter ellipsoids
were drawn at the 30% probability level.

478
M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ485
/
attributed to additional donation from an occupied
oxygen orbital to a low-lying empty metal orbital [8].
The phenyl ring is nearly parallel to the plane defined by
the C(8), C(9), and O(3) atoms, in agreement with a high
degree of conjugation in the coordinated acetophenone
molecule.
In the present case two n(CO) bands were expected
for each isomer. The observation of only three bands
indicates accidental overlap of two bands (one for each
isomer): indeed in cyclohexane the two components of
the band at intermediate wavenumbers (the most intense
one, at 1909 cmꢂ1 in Et2O) are partially resolved. On
the base of the above considerations, we attribute to the
p isomer the n(CO) band at the highest wavenumbers
(1979 cmꢂ1 in Et2O) and to the s form the band at the
lowest wavenumbers (1851 cmꢂ1). Theses attributions
lead to a close agreement with data concerning related
species: see for instance n(CO) 1992, 1912 cmꢂ1 (in
Et2O) for [Cp(CO)2Re(h2-CH2O)] [13] and 1921, 1850
cmꢂ1 for [Cp(CO)2Re(OEt2)] [14], that contains the
labile ligand Et2O, (s-donor through an oxygen atom),
comparable with acetophenone.
2.2. Solution behavior of 4. The sꢀp equilibrium
/
The IR spectrum of solid 4 (nujol mull) shows two
main n(CO) absorptions (1905s, 1890sh and 1817s
cmꢂ1) and a weak band at 1557 cmꢂ1, due to the
ketonic carbonyltesi p.75, in agreement with the solid
state structure. The IR spectrum in solution, however, is
markedly different, because it shows three main bands,
whose relative intensities vary on changing the solvent
(see Table 3). Moreover, one band is at wavenumbers
much higher (1979 cmꢂ1 in Et2O) than those observed
in solid. This indicates the presence in solution of two
species (whose relative ratio varies with the solvent),
that can be identified as two isomers differing for the
binding mode of the ketone molecule. Indeed, it is well
known [9] that aldehyde or ketone ligands can adopt
either a h1 (s) or a h2 (p) coordination of the carbonylic
group, and often the solutions of their complexes
contain equilibrium mixtures of the two isomers. The
h2-coordination involves substantial back-bonding from
the metal to a vacant p* orbital of the organic carbonyl
[10], that is reflected in the dramatic decrease of the
wavenumbers both of the n(CO) IR absorption (below
The relative intensity of the lowest wavenumbers
band increases on passing from n-hexane to Et2O to
THF or acetone (see Table 3), indicating that polar
solvents favor the s form. An opposite trend was
previously observed by Gladysz for some benzaldehyde
cationic complexes [12a,12b] and was attributed to the
fact that in the p isomers the positive charge remains
more localized than in the s-isomers, resulting in
enhanced solvation energy. In our neutral species, it is
likely that the s-isomer acquires a higher polarity than
the p-isomer due to the more classical donorꢀ
/
acceptor
character of the ketone-metal s-interaction with respect
to the synergic donation/back-bonding interaction op-
erative in the p-isomer.
Variable temperature IR spectra (both in n-hexane
and in CH2Cl2) showed a slight variation of the relative
intensities of the resonances at the highest (p) and lowest
(s) frequency, indicative of the increase of the p/s ratio
increases as the temperature decreases. This trend was
previously observed in all the cases in which such type of
equilibria was investigated [11,12,15].
The variable temperature 1H and 13C NMR data
(Figs. 3 and 4) provided further insight into this
equilibrium. In CD2Cl2, at 183 K, two sets of signals
were observed (see Table 4), in the ratio approximately
1:2.7. The minor species at this temperature had sharp
resonances both in the 1H and 13C spectra, and has been
identified as the s-isomer, on the bases of the position of
the 13C resonance of the carbonyl of coordinated
acetophenone (211.3 ppm, in the ratio 1:2 vs. the unique
signal of the two metal carbonyls, at 205.7 ppm) [16].
The corresponding resonance for the p-isomer was
observed at 83.6 ppm: similar upfield values (range
1200 cmꢂ1) and of the 13C NMR resonance (below dꢁ
/
100 ppm) for the ketonic carbonyl [9]. Such back-
bonding is less important in s-complexes and accord-
ingly IR and 13C NMR data for these carbonyls are
much closer to the values of the free organic ligands. In
the p-form, therefore, competition for back-bonding
from the metal d-orbitals between the organic and the
metal carbonyls causes the increase of the wavenumbers
of the n(CO) modes of the metal-carbonyls, with respect
to the analogous modes in the complexes where the
ketone is s-bound [11]. Similar arguments have been
applied to the n(NO) mode in the numerous
[CpRe(NO)(PPh3)L]ꢃ complexes (Lꢁ
aldehyde or ke-
/
tone) [12].
Table 3
Ratio between the absorbances of the nCO IR bands of 4 at the highest
and at the lowest wavenumbers (due to the p and s form, respectively),
in different solvents at room temperature
60ꢀ90 ppm) are typical of p-complexes of aldheydes and
/
ketones [9].
Solvent
p/s ratio
The unambiguous attribution of each signal (1H and
13C) of the phenyl group could not be achieved for the
extensive overlap of the resonances of the two isomers
and the uncompleted averaging of the unequivalent
phenylic resonances of two p-isomers. Indeed, at 183 K
n-Hexane
Et2O
THF
3.6
2.3
0.90
0.85
Acetone

M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ
/485
479
NMR resonances [7,18]. In the present case, the sharp-
ness of the signals of the s form even at the lowest
temperature (183 K) could be due to a particularly fast
E/Z exchange rate. Indeed, it has been previously
observed [18] that in d6 Re complexes the presence of
two filled degenerate orbital available for back-bonding
leads to a low activation barrier and the two isomers are
not distinguishable under conventional NMR condi-
tions [7]. However, the lack of IR evidence for two
isomers prompted us to assume that here only the E
isomer has a significant concentration in solution,
because steric hindrance between the carbonyls and
the hydrogens of the phenyl ring destabilizes the Z
isomer.
Increasing the temperature, the exchange between s
and p-isomers became faster, leading to a generalized
broadening of the signals and eventually to averaged
resonances (Figs. 3 and 4). The positions of these
averaged resonances indicate that, as expected, the p/s
ratio decreases on increasing the temperature, from 2.9
to approximately 1. Particularly reliable for this estimate
is the resonance of ketonic carbonyl (averaged signal at
d 142.5 ppm at 263 K), since the high Dd of the
resonances of the s and p forms (Ddꢁ
183 K) makes less significant the temperature depen-
dence of their chemical shifts. A DH8 value of ꢂ4.5 kJ
molꢂ1 was estimated for the sU
p reaction, on using
the values of Kꢁp/s evaluated at 183 K (from the
integrated intensities of the separated resonances of the
two isomers, averaged values from different signals, Kꢁ
2.87) and at 263 K (from the chemical shift of the molar-
fraction-weighted average resonance, Kꢁ1.16). Values
in the range 7ꢀ
14 kJ molꢂ1 have been estimated for
/
127.7 ppm, at
/
/
/
/
Fig. 3. 1H NMR variable temperature spectra of compound 4 (in
CD2Cl2, 11 T). Traces have been plotted with different vertical scales.
The upper trace refers to the same solution after 40 h at 295 K in
CD2Cl2: * acetophenone; m [CpRe(CO)3]. The inset shows the
diastereotopic hydrogen atoms of 5 observed when decomposition
occurred in CH2Cl2.
/
/
related equilibria involving aldehyedes [12b].
Fig. 5 shows the Arrhenius plot of the rate constants
for the pUs interconversion at different temperatures
/
evaluated by band shape analysis of the methyl reso-
1
nances (both in the H and 13C spectra) and of the 13C
all the 1H and 13C resonances of the p-isomer were
broad, indicating the occurrence of an exchange process
between the two different p forms depicted in Scheme 2.
Similar processes have been previously recognized for p-
bonded aldehydes [12b,17], and in the case of [CpRe-
signals of the Cp groups. The different evaluations are in
satisfactory agreement and an activation energy of 58(1)
kJ molꢂ1 was accordingly obtained.
2.3. Decomposition of 4 in CH2Cl2
(NO)(PPh3)(h2-OC(H)Ar]ꢃ (Arꢁ
was proved that the pꢀp? interconversion occurred with
/
p-C6H4OCH3) [17] it
/
IR and NMR monitoring of the room temperature
behavior of 4 in dichloromethane showed quite fast
decomposition (complete in a few hours) to give free
acetophenone, some [CpRe(CO)3] and the novel com-
plex [Cp(CO)2Re(Cl)(CH2Cl)] (5), analogous to the
the intermediacy of the s-isomer. In the present case, on
the contrary, the signals of the s-isomer are sharp, and,
therefore, this isomer cannot be involved in the inter-
conversion of the two p forms. This also implies that the
pꢀ
/
p? interconversion is an intramolecular process,
previously known pentamethylꢀcyclopentadienyl deri-
/
because dissociation would offer the chance of pꢀ
/
s
vative [Cp*(CO)2Re(Cl)(CH2Cl)] [19]. The latter species
was obtained from oxidative addition of CH2Cl2 to the
unsaturated fragment Cp*(CO)2Re, originated by THF
loss from [Cp*(CO)2Re(THF)]: most likely this is also
the route responsible for the formation of 5 from the
ketonic complex 4 (Eq. (4)). Interestingly, the addition
interconversion and, therefore, would broaden also the
signal of the s-isomer.
In several related ketone complexes, the occurrence of
interconversion between the Z and E isomers of the s
form was revealed by low temperature broadening of the

480
M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ485
/
Fig. 4. Selected regions of 13C NMR variable temperature spectra of 4: (left) carbonylic (ꢄ
(CD2Cl2, 11 T). Inset: averaged signal due to the carbonyl of the coordinated acetophenone. The upper trace refers to the same solution after 40 h at
295 K in CD2Cl2: * acetophenone; “ [CpRe(CO)3].
/4), (middle) aromatic and (right) aliphatic regions
/
to 4 of acetophenone (10 equiv.) did not change the
reaction course, indicating that acetophenone dissocia-
tion is irreversible in CH2Cl2.
spectrum (obviously not detected when the decomposi-
tion of 4 occurred in CD2Cl2, top trace in Fig. 3). As
expected, the n(CO) bands of 5 are shifted to higher
wavenumbers (20 cmꢂ1) with respect to those of the
Cp* derivative [19]. The simultaneous formation of
[CpRe(CO)3] is attributable to some decomposition,
causing evolution of CO, which is captured by the
Cp(CO)2Re fragment.
[Cp(CO)2RefOC(Me)Phg]ꢃCH2Cl2
0 [Cp(CO)2Re(Cl)(CH2Cl)]ꢃPh(Me)CO
(4)
Particularly diagnostic for the formulation of 5 is the
AB pattern of the CH2Cl group, in the 1H NMR

M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ
/485
481
Table 4
1H and 13C NMR data of 4s and 4p in CD2Cl2 at 183 K (d, ppm, 11
T)
if only at high temperature) [5], and a similar reaction
has been suggested as the step responsible for the
formation of free ketones in the alkylation of iron
acylates by different RX reactants [21]. Moreover, a
benzophenone complex has been prepared by thermo-
lyzing a phenyl-h2-benzoyl zirconocene [22].
Isomer
1H
13C
Me
Me
Cp
Cp
COketone
COmetal
s
p
2.63 5.32 27.6 84.9 211.3
2.22 5.09 29.1 92.4 83.6
205.7
202.6, 207.9
[Cp(CO)2Re(Me)fC(O)Phg]
0 [Cp(CO)2RefOC(Me)Phg]
(5)
However, this assumption seems not to be valid.
Spectroscopic pursuit of the alkylation reaction clearly
shows that the formation of 4 is parallel (and not
consecutive) to the formation of 3. Moreover, com-
pound 3 is stable in diethyl ether solution at room
temperature for a time much longer than that of the
above reaction and anyway its decomposition does not
afford any 4. Finally, we have observed that the amount
of 4 in the reaction mixture does not increase with time
(on the contrary, it tends to decrease, due to the
instability of 4 in solution).
Scheme 2.
The formation of the ketone complex seems to be
restricted to the use of Et2O as solvent, methyltriflate as
alkylating agent and 1 as substrate. Indeed changes of
the solvent (on using either cyclohexane, or THF or
acetone) and of the alkylating agent (MeI, Me3OBF4)
dramatically reduced the amount of 4 (with the partial
exception of cyclohexane, as described in Section 3). The
alkylation of the acylmetallate Li[Cp(CO)2ReC(O)Me]
(with methyltriflate in Et2O) gave the same products
observed by Bergman [5], without any acetone complex.
The peculiarity of the reaction led us to investigate on
the possible involvement of radical species (that might
account also for some variability in the relative yields of
the three main products in repeated preparations) [23].
However, no evidence for this has been so far obtained.
We have observed that 4 is formed also in the dark, and
that the use of freshly distilled or ‘old’ ethyl ether, as
well as the addition of a radical scavenger (TEMPO)
does not cause any significant change. ESR, also in the
presence of a spin trap (2-methyl-2-nitroso-propane),
did not give any clear evidence of radicals.
Fig. 5. Arrhenius plot of the rate constants for the pls evaluated
/
from the band shape analysis of the H Me (k), 13C Me (2) and 13C
1
With all this evidence, the most reasonable explana-
tion for the formation of 4 is direct methylation at the
carbon atom of the acylate group, in an h2 canonical
form in which the negative charge is localized just on the
carbon atom (Scheme 3) [24]. This form should acquire
some stability, with respect to the more common
canonical forms having the charge localized on the
oxygen or the metal atoms, from the well-known
oxophilicity of rhenium. Indeed the h2-coordination of
acyl groups is quite common among the oxophilic early-
transition-metals [25], and several examples of this
coordination are known also for transition metals of
Cp (\) resonances, respectively.
2.4. Suggestions for the mechanism of the reaction
affording 4
The formation of a ketone complex in the alkylation
of an acylmetallate anion is unprecedented. The most
reasonable route to 4 would be the process depicted in
Eq. (5), involving reductive elimination of acetophenone
from the benzoylꢀmethyl complex 3, followed by
/
acetophenone coordination [20]. Ketone reductive elim-
ination has been previously observed for the analogous
Groups 5ꢀ
/
8 [26]. Therefore, we might speculate that
acetylꢀmethyl complex studied by Bergman et al. (even
/
only CF3SO3Me affords the ketone complex 4, because

482
M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ485
/
second fractions contained mainly CpRe(CO)3. The first
fraction (yellow) contained mainly 2 (which was isolated
by precipitation from n-hexane at 193 K, yields 30% ca.)
and other species, whose separation required a second
chromatographic separation (silica gel 2.5ꢄ15 cm;
/
eluent n-hexane). This second chromatography af-
forded, as first colorless fraction, a non-carbonyl
compound, identified as (C6H5)2 (biphenyl) by compar-
ison with literature data [28], as second colorless fraction
the species [Cp(CO)2ReMe2], identified from its IR and
NMR data [5] and as third fraction the yellow com-
pound 2. The third fraction of the first separation (light
brown) was constituted by 3 (yield ca. 10% after
Scheme 3.
it is the most powerful (and, therefore, less selective)
among the tested alkylating agents (the other, weaker,
electrophyles select the methylation site on the bases of
their affinity). The h2-intermediate could be stabilized
by the presence of the phenyl substituent (able to
delocalize the negative charge on the carbon atom),
and this could explain the lack of alkylation at the
carbon atom in the case of the methyl derivative. More
difficult to explain is the role of the solvent, since either
more polar or less polar solvents, with respect to diethyl
ether, contrast the formation of 4. Further investigation
on this and similar systems are therefore, required.
crystallization from diethyl etherꢀn-hexane at 193 K).
/
The second (red) fraction, after evaporation to dryness,
afforded a residue that was crystallized from n-hexane
at 193 K, yielding 4 (10% ca.). Crystals of 4 suitable for
X-ray diffraction were obtained from diethyl etherꢀn-
/
hexane at 243 K. Anal. Calc. for C15H13O3Re: C, 42.14;
H, 3.06. Found: 2 C, 42.3; H, 3.2%; 3 C, 42.4; H, 3.3%; 4
1
C, 42.1; H, 3.1%: Spectroscopic data: IR and H NMR
see Table 1; 13C NMR (CD2Cl2, 295 K) 3, d 225.17
3. Experimental
(ReÃ
128.20, 127.01 (C ortho and meta), 109.2 (C ipso), 94.24
Cp), ꢂ CH3); 4 d 204.18, 142.5, 130.17,
/
C(O)Ph), 200.11 (ReÃ/CO), 130.59 (Ph, C para),
The reactions were performed under nitrogen, using
the Schlenk technique, and solvents deoxygenated and
dried by standard methods. Literature methods were
used for the preparation of CpRe(CO)3 [27],
Li[Cp(CO)2ReC(O)Ph] (Liꢃ1) [2], and Li[Cp(CO)2Re-
C(O)Me] [5]. IR spectra were obtained on a Bruker
Vector 22 FT instrument using a 0.1 mm CaF2 cell. Low
temperature IR spectra were recording using a variable
temperature cell (Specac P/N 21425). H and 13C NMR
spectra were recorded on Bruker AC200 and AMX500
spectrometers. The calibration of the temperature has
(ReÃ
/
/
35.52(ReÃ
/
128.73, 126.40, 88.10, 28.06; see Table 4 for the low
temperature spectra of the two isomers of 4.
3.1.2. Reaction at 273 K
A sample of Liꢃ1 (87 mg, 0.21 mmol) was suspended
in diethyl ether and the reaction flask was kept at 273 K.
The suspension was treated with CF3SO3Me (28 ml, 0.24
mmol) and stirred. IR monitoring showed completion of
the reaction in 24 h approximately, to give a mixture
constituted mainly by compounds 2, 3, and 4, in a ratio
comparable with that of the same reaction at r.t..
1
been checked using the standard CD3ODꢀ
solution. Mass spectra were recorded on a VG
7070EQ instrument, in the FABꢃ mode, using NBA
as matrix.
/
CH3OH
/
3.1.3. Reaction at 308 K
A sample of Liꢃ1 (170 mg, 0.40 mmol) suspended in
diethyl ether and treated with CF3SO3Me (56 ml, 0.49
mmol), was refluxed on a water bath at 310 K until IR
monitoring showed completion of the reaction (75 min),
to give a mixture of 2, 3, and 4, in a ratio substantially
comparable with that observed at r.t..
3.1. Reaction of Liꢃ[Cp(CO)2ReC(O)Ph]ꢂ (Liꢃ1)
with CF3SO3Me in diethyl ether
3.1.1. Reaction at room temperature, on a preparative
scale
A 100 ml Schlenk flask was charged with Liꢃ1 (465
mg, 1.10 mmol), a stir bar and diethyl ether (35 ml). The
suspension was treated at room temperature (r.t.) with a
slight excess of CF3SO3Me (150 ml, 1.32 mmol) and the
mixture was stirred for 2 h approximately. In this time
the color changed from yellow to deep red. The
suspension was concentrated under vacuum, and sepa-
rated by chromatography on silica gel with 3:1 (v/v) n-
3.2. Reaction of Liꢃ[Cp(CO)2ReC(O)Ph]ꢂ (Liꢃ1)
with CF3SO3Me in acetone
A sample of Liꢃ1 (300 mg, 0.71 mmol) was
suspended in 20 ml of acetone and treated at r.t. with
CF3SO3Me (90 ml, 0.80 mmol). The mixture was stirred
for 1.5 h. During the reaction the color changed from
yellow to orange. IR monitoring revealed that 2 was
largely dominant in the reaction mixture. Formation of
3 or 4 was not detected. Chromatographic separation on
silica gel with 4:1 (v/v) n-hexane: diethyl ether afforded a
hexane: diethyl ether (16ꢄ2.5 cm), affording three main
/
fractions (yellow, red and light brown, respectively). A
colorless fraction collected between the first and the

M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ
/485
483
dark orange fraction: the IR spectrum (n-hexane)
showed the presence of 2 and of another unknown
species, responsible for n(CO) bands at 1983sh, 1978s,
1907s cmꢂ1 and an 1H resonance at d 4.80 ppm (C6D6).
3.8. Reaction of Li[Cp(CO)2ReC(O)Me] with
CF3SO3Me in diethyl ether
A sample of CpRe(CO)3 (50 mg, 0.15 mmol) in 5 ml
of diethyl ether was treated with a 1.6 M CH3Li/THF-
cumene solution (120 ml, ca. 0.19 mmol). After 2 h, the
precipitate of Li[Cp(CO)2ReC(O)Me] [5] was isolated,
dried, suspended in diethyl ether and treated at r.t. with
CF3SO3Me (16 ml, 0.14 mmol). After 2 h IR monitoring
showed the formation mainly of [Cp(CO)2R-
e(Me)(COMe)], accompanied by a smaller amount (ca.
20%) of [Cp(CO)2Re{C(OMe)Me}].
3.3. Reaction of Liꢃ[Cp(CO)2ReC(O)Ph]ꢂ (Liꢃ1)
with CF3SO3Me in THF
A sample of Liꢃ1 (15 mg, 0.036 mmol), was dissolved
in THF (4 ml) and treated at 193 K with CF3SO3Me (5
ml, 0.044 mmol). The solution was warmed up to r.t. and
the color turned from yellow to orange. IR and NMR
revealed the formation of 2 and 3 (ratio 2:1 ca.) and of a
very small amount of 4 (ca. 5% of the reaction mixture).
3.9. Reaction of 4 with CH2Cl2
A sample of 4 (10 mg, 0.024 mmol) was dissolved at
r.t. in CH2Cl2 (3 ml ca.). IR monitoring showed the
progressive decrease of the intensity of the carbonyl
bands of 4, substituted by new bands, attributable to
[CpRe(CO)3] (2024 and 1926 cmꢂ1), free acetophenone
(1685 cmꢂ1), and to the novel species [Cp(CO)2R-
eCl(CH2Cl)] (5) (at 2042s and 1959vs cmꢂ1). After 5 h
the reaction went to completion. The NMR spectrum
(CD2Cl2, r.t.) showed the resonances of free acetophe-
3.4. Reaction of Liꢃ[Cp(CO)2ReC(O)Ph]ꢂ (Liꢃ1)
with CF3SO3Me in cyclohexane
A sample of Liꢃ1 (45 mg, 0.11 mmol), was suspended
in cyclohexane (8 ml) and treated at r.t. with CF3SO3Me
(13 ml, 0.11 mmol). The mixture was stirred at r.t..
During the reaction time the color of the solution
became pale red. The IR spectrum recorded after 2 h
showed the formation mainly of 2 and 3 (ratio ca. 1:1),
with minor amounts of 4 (ca. 10%).
none (d 7.90ꢂ
ppm), and of 5, at d 5.85 (s, 5H), 4.72 (d, 1H), 4.49 (d,
1H, JHÃH 9 Hz) ppm.
/
7.37, 2.51 ppm), of [CpRe(CO)3] (d 5.34
ꢁ
/
3.5. Reaction of Liꢃ[Cp(CO)2ReC(O)Ph]ꢂ with
(CH3)3OBF4 in diethyl ether
3.10. X-ray data collection and structure refinement
A sample of Liꢃ1 (56 mg, 0.13 mmol) suspended in
diethyl ether (6 ml) was treated at r.t. with (CH3)3OBF4
(23 mg, 0.15 mmol). The IR spectrum after 2.5 h showed
the formation of 2 and 3 (ratio ca. 4:1) which were
isolated by chromatography, as described above.
A suitable crystal of 4 was mounted in air on a glass
fibre tip onto a goniometer head. Single-crystal X-ray
diffraction data were collected on an Enrafꢀ
/
Nonius
CAD4 diffractometer using graphite-monochromated
˚
0.71073 A) at r.t.. Unit cell
parameters were obtained from least-squares fitting of
the setting angles of 25 randomly distributed intense
Mo Ka radiation (lꢁ
/
3.6. Reaction of Liꢃ[Cp(CO)2ReC(O)Ph]ꢂ with
CH3I in diethyl ether
reflections with 105
/
u 5148. Data collection was per-
/
formed by the v scan method with variable scan speed
(maximum time per reflection 70 s) and variable scan
A sample of Liꢃ1 (33 mg, 0.079 mmol), suspended in
diethyl ether (3 ml) was treated at r.t. with CH3I freshly
distilled (14 ml, 0.22 mmol) and stirred at r.t. for 3 days,
protected from light. IR monitoring showed formation
only of compound 3.
ranges (1.00ꢃ0.35 tan u (8)). Crystal stability under
/
diffraction was checked by monitoring three standard
reflections every 180 min. The measured intensities were
corrected for Lorentz, polarization, decay, and back-
ground effects and reduced to Fo2. An empirical
absorption correction was applied using c scans of
three suitable reflections having x values close to 908
[29]. Crystal data and data collection parameters are
summarized in Table 5.
3.7. Parallel alkylation reactions in the dark and in the
light
Two Schlenk tubes were charged with the same
amount of Liꢃ1 (40 mg ca., 0.095 mmol), and
suspended in freshly distilled diethyl ether (5 ml). One
of the Schlenk tube was protected from light. Both the
tubes were treated at r.t. with CF3SO3Me (12 ml ca., 0.10
mmol). The IR spectra recorded after stirring the
mixtures for 2 h did not show any significant difference.
The structures were solved by direct methods [30] and
subsequent Fourier synthesis; they were refined by full-
matrix least-squares on F2 using reflections with I ꢀ
/
2s(I) [31]. Scattering factors for neutral atoms and
anomalous dispersion corrections were taken from the
internal library of SHELX97. Weights were assigned to
individual observations according to the formula wꢁ1/
/

484
M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ485
/
Table 5
Summary of crystal data and structure refinement parameters for 4
observed and calculated structure factors moduli for 4
(8 pages) are available from the authors on request.
Formula
C15H13O3Re
Formula weight
Crystal system
Space group
427.45
References
monoclinic
P21/c (No. 14)
12.172(3)
8.338(2)
14.856(2)
112.46(2)
1393.4(5)
4
[1] (a) G. Ciani, M. Moret, A. Sironi, P. Antognazza, T. Beringhelli,
G. D’Alfonso, R. Della Pergola, A. Minoja, J. Chem. Soc., Chem.
Commun. 18 (1991) 1255;
˚
a (A)
˚
b (A)
˚
c (A)
(b) P. Antognazza, T. Beringhelli, G. D’Alfonso, A. Minoja, G.
Ciani, M. Moret, A. Sironi, Organometallics 11 (1992) 1777;
(c) T. Beringhelli, A. Ceriotti, G. Ciani, G. D’Alfonso, L.
Garlaschelli, R. Della Pergola, M. Moret, A. Sironi, J. Chem.
Soc., Dalton. Trans. 1 (1993) 199;
b (8)
3
V (A )
˚
Z
F(000)
Density (g cmꢂ3
Absorption coefficient (mmꢂ1
Crystal description
Crystal size (mm)
u Range (8)
808
2.038
)
(d) T. Beringhelli, G. D’Alfonso, A. Minoja, Organometallics 13
(1994) 663;
)
8.721
red prism
(e) M. Bergamo, T. Beringhelli, G. Ciani, G. D’Alfonso, M.
Moret, A. Sironi, Organometallics 15 (1996) 1637;
(f) M. Bergamo, T. Beringhelli, G. Ciani, M. Moret, A. Sironi,
Inorg. Chim. Acta 259 (1997) 291;
0.26ꢄ
3.05u5
145h 5
l 517
/
0.21ꢄ
25.0
13, 05
/
0.14
/
/
Index ranges
ꢂ/
/
/
/
k 5
/
9, 05
/
/
(g) M. Bergamo, T. Beringhelli, P. Mercandelli, M. Moret, A.
Intensity decay (%)
Transmission factors (min, max)
Collected reflections
5
Sironi, Inorg. Chim. Acta 300ꢀ302 (2000) 1022.
[2] E.O. Fischer, A. Riedel, Chem. Ber. 101 (1968) 156.
[3] F.G.A. Stone, Angew. Chem., Int. Ed. Engl. 23 (1984) 89.
/
0.186, 0.375
2437
Reflections with I ꢀ
/
2s(I)
2077
[4] H. Fischer, in: K.H. Dots, H. Fischer, P. Hofmann, F.R. Kreissl,
¨
U. Schubert, K. Weiss (Eds.), Transition Metal Carbene Com-
plexes, VCH, Weinheim, 1983, p. 1.
Data/parameters
Weights (a, b) a
2077/173
0.04, 1.17
1.115
Goodness-of-fit S(F2) b
R(F) c
[5] I.K. Goldberg, R.G. Bergman, J. Am. Chem. Soc. 111 (1989)
1285.
0.0243
0.0611
WR(F2) d
[6] Y. Tanimoto, H. Kobayashi, S. Nagakura, Y. Saito, Acta
Crystallogr., Sect. B 29 (1973) 1822.
Maximum difference peak, hole (e 0.758, ꢂ0.917
/
ꢂ3
˚
A
)
[7] M.D. Dalton, J.M. Ferna´ndez, K. Emerson, R.D. Larsen, A.M.
Arif, J.A. Gladysz, J. Am. Chem. Soc. 112 (1990) 9198.
[8] B. Galeffi, M. Simard, J.D. Wuest, Inorg. Chem. 29 (1990) 951.
[9] Y.H. Huang, J.A. Gladysz, J. Chem. Educ. 65 (1988) 298.
[10] F. Delbecq, P. Sautet, J. Am. Chem. Soc. 114 (1992) 2446.
[11] See for instance: (a) D.M.Schuster, P.S. White, J.L. Templeton,
Organometallics 19 (2000) 1540; (b) D.M. Schuster, P.S. White,
J.L. Templeton, Organometallics 15 (1996) 5467.
a
wꢁ
/
1/[s2(Fo2)ꢃ
[aw(Fo2ꢂFc2)2/(nꢂ
and p is the number of refined parameters.
/
(aP)2ꢃ
/
bP], where Pꢁ
/
(Fo2ꢃ
/
2Fc2)/3.
b
Sꢁ
/
/
/
p)]1/2, where n is the number of reflections
c
R(F)ꢁ
/
ajjFojꢂ
/
jFcjj/ajFoj.
d
wR(F2)ꢁ[aw(Fo2ꢂ
/
/
Fc2)2/awFo4]1/2
.
[s2(Fo2)ꢃ
/
(aP)2ꢃ
/
bP], where Pꢁ
/
(Fo2ꢃ2Fc2)/3; a and b
/
[12] (a) N. Quiros Mendez, A.M. Arif, J.A. Gladysz, Angew. Chem.,
Int. Ed. Engl. 29 (1990) 1473;
were chosen to give a flat analysis of variance in terms of
Fo2. Anisotropic displacement parameters were assigned
to all non-hydrogen atoms. Hydrogen atoms were
placed in idealized position and refined riding on their
parent atom with an isotropic displacement parameter
1.2 times that of the pertinent carbon atom. The
hydrogen atoms of the methyl group has been modeled
as disordered over two positions (with an occupation
factor of 1/2 each) rotated from each other by 608. Final
difference electron density map showed no features of
chemical significance, with the largest peaks lying close
to the metal atom. Final conventional agreement
indexes and other structure refinement parameters are
listed in Table 5.
(b) N. Quiro`s Me´ndez, J.W. Sayler, J.A. Gladysz, J. Am. Chem.
Soc. 115 (1993) 2323.
[13] H. Berke, R. Birk, G. Huttner, L.Z. Zsolnai, Naturforsch. 39b
(1984) 1380.
[14] We measured these IR data on the sample obtained by low
temperature UV irradiation of CpRe(CO)3 in Et2O, as described
in Ref. [13].
[15] D.P. Klein, D.M. Dalton, N.Q. Mendez, A.M. Arif, J.A. Gladysz,
J. Organomet. Chem. 412 (1991) C7.
[16] In the related complex [CpRe(NO)(PPh3)(OC(Me)Ph)]ꢃ, a value
of 215.9 was observed for the s-bound acetophenone carbonyl [7].
[17] (a) N. Quiros Mendez, C.L. Mayne, J.A. Gladysz, Angew. Chem.,
Int. Ed. Engl. 29 (1990) 1475;
(b) B. Boone, D.P. Klein, J.W. Seyler, N. Quiro`s Me´ndez, A.M.
Arif, J.A. Gladysz, J. Am. Chem. Soc. 118 (1996) 2411.
[18] See J.R. Caldarelli, L.E. Wagner, P.S. White, J.L. Templeton, J.
Am. Chem. Soc. 116 (1994) 2878 and refs. therein.
[19] O.J. Scherer, M. Ehses, G. Wolmershauser, Z. Naturforsch 52b
¨
(1997) 762.
[20] Reductive elimination in Et2O solution would afford the solvent
stabilized unsaturated intermediate [Cp(CO)2Re(OEt2)]. We have
verified that the addition of acetophenone to a solution of this
species (obtained by low temperature UV irradiation of
[CpRe(CO)3]) affords 4 (even if an excess of acetophenone is
required to make quantitative the reaction).
4. Supplementary materials
A complete list of atomic coordinates, anisotropic
displacement parameters, bond lengths and angles,

M. Bergamo et al. / Inorganica Chimica Acta 350 (2003) 475ꢀ
/485
485
[21] (a) M.F. Semmelhack, R. Tamura, J. Am. Chem. Soc. 105 (1983)
4099;
(b) J.P. Collman, Acc. Chem. Res. 8 (1975) 342.
[25] See for instance K. Tatsumi, A. Nakamura, P. Hofmann, P.
Stauffert, R. Hoffmann, J. Am. Chem. Soc. 107 (1985) 4440, and
references therein.
[22] F. Rosenfeldt, G. Erker, Tetrahedron Lett. 21 (1980) 1637.
[23] We have evaluated the relative intensities of the nCO IR bands of
the three main reaction products, in the reaction mixtures
obtained in 19 different alkylation reactions with MeOTf, in
Et2O, at room temperature or at 273 K, in the light or in the dark.
The average intensities of the absorptions of 3 and 4 (at 1940 and
1979 cmꢂ1, respectively) relative to the intensity of the IR band of
2 at 1958 cmꢂ1 were 0.96 and 0.53, with standard deviation 0.21
and 0.09, respectively. No correlation was observed between these
relative intensities and the reaction temperature or the presence/
absence of light.
[26] See for instance (a) M.D. Curtis, K.-B. Shiu, W.M. Butler, J. Am.
Chem. Soc. 108 (1986) 1550; (b) C. Jablonski, G. Bellachioma, G.
Cardaci, G. Reichenbach, J. Am. Chem. Soc. 112 (1990)
1632.
[27] R.B. King, R.H. Reimann, Inorg. Chem. 15 (1976) 179.
[28] M. Yanagisawa, K. Hayamiz, O. Yamamoto, Magn. Reson.
Chem. 25 (1987) 184.
[29] A.C.T. North, D.C. Phillips, F.S. Mathews, Acta Crystallogr., A
24 (1968) 351.
[30] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M.C.
Burla, G. Polidori, M. Camalli, J. Appl. Crystallogr. 27 (1994)
435.
[24] I. Tkatchenko, in: G. Wilkinson, F.D.A. Stone, E.W. Abel (Eds.),
Comprehensive Organometallic Chemistry, vol. 8, I ed, Pergamon
Press, Oxford, 1982, p. 101.
[31] G.M. Sheldrick, SHELX97, Universitat Gottingen, Gottingen,
¨ ¨ ¨
Germany, 1997.